A CNN Centric Approach to Acute Lymphoblastic Luekemia Detection

Leukaemia is a serious malignancy that impacts both adults and children, presenting as either a chronic or acute condition. Acute Lymphoblastic Leukemia (ALL) stands as the most prevalent type of blood disease and remains a leading cause of death. This aggressive cancer invades the blood, spreading via surrounding bodily systems and organs. The conventional method for classifying cancer cells and non-cancer cells involves manual diagnosis, where specialists inspect cell images under a microscope and annotate the labels. However, this manual microscopic analysis is not only tedious but also susceptible to potential false diagnoses. In response to these challenges, this project seeks to leverage a pre-trained model for automated de-tection and classification of leukemia subtypes, including benign, early, pre, and pro stages. By utilizing a pretrained model, the proposed system aims enhance its ability to accurately discern and classify leukemia subtypes. By leveraging pre-trained models, the proposed approach aims to improve classification accuracy, efficiency, and reduce the risk of misdiagnosis compared to manual analysis. This research contributes to the advancement of automated leukemia diagnosis, paving the way for potential clinical applications.